An independent search for Jovian neutrinos using BOREXINO data

TL;DR

The paper independently reproduces the Jovian neutrino signal reported in A24 using BOREXINO $^7$Be data, applying Bayesian regression and model comparison to rate-time-series and monthly modulation data. While a Jovian flux of about $6\%$ of the solar signal emerges with ~$2\sigma$ significance in some epochs, Bayesian evidences (Bayes factors $B_{21}\lesssim 5$) are marginal, and full-data analyses reduce the signal strength with Bayes factors near unity. The study also reveals spurious signals from Venus and Saturn under certain priors, though model comparison disfavors persistent planetary contributions beyond the Sun; frequentist analyses yield consistent, but also inconclusive, results. Overall, the results do not provide robust evidence for Jovian neutrinos in the full data set, underscoring the need for additional cross-checks and independent measurements to confirm or refute the Jovian neutrino hypothesis.

Abstract

In a recent study, arXiv:2401.13043 found evidence for a 6% flux contribution from Jupiter to the total flux rate time series data from the BOREXINO solar neutrino experiment, specifically during the time intervals 2019-2021 and 2011-2013. The significance of this detection was estimated to be around $2σ$. We reanalyze the BOREXINO data and independently confirm the Jovian signal with the same amplitude and significance as that obtained in arXiv:2401.13043. However, using the same technique, we also find a spurious flux contribution from Venus and Saturn (at $\sim 2σ$ significance), whereas prima facie one should not expect any signal from any other planet. We then implement Bayesian model comparison to ascertain whether the BOREXINO data contain an additional contribution from Jupiter, Venus or Saturn. We find Bayes factors of less than five for an additional contribution from Jupiter, and less than or close to one for Venus and Saturn. This implies that the evidence for an additional contribution from Jupiter is very marginal.

Figures (20)

• Figure 1: Marginalized 68% and 95% credible intervals for $\mathcal{R}_{\mathrm{B}}$ and $\mathcal{R}_{\mathrm{jup}}$ after using uniform prior on $\frac{\mathcal{R}_{\mathrm{jup}}}{(5\,\mathrm{AU})^2} \in \mathcal{U} [0, 4]$ and normal prior on $\mathcal{R}_{\mathrm{sun}} \in \mathcal{N} (25,2)$ with units of (cpd/100t). For this plot, we have used DE442s ephemeris. The marginalized 68% value for $\frac{\mathcal{R}_{\mathrm{jup}}}{(5\,\mathrm{AU})^2}$ is given by $1.53^{+0.78}_{-0.75}$ (cpd/100t).
• Figure 2: Marginalized 68% and 95% credible intervals for $\mathcal{R}_{\mathrm{sun}}$, $\mathcal{R}_{\mathrm{jup}}$ and $\mathcal{R}_{\mathrm{B}}$ after using uniform prior on $\frac{\mathcal{R}_{\mathrm{jup}}}{(5\,\mathrm{AU})^2} \in \mathcal{U} [0, 4]$ and uniform prior on $\mathcal{R}_{\mathrm{sun}} \in \mathcal{U} [0, 50]$ with units of (cpd/100t). For this plot, we have used DE442s ephemeris. The marginalized 68% value for $\frac{\mathcal{R}_{\mathrm{jup}}}{(5\,\mathrm{AU})^2}$ is given by $1.31^{+0.84}_{-0.77}$ (cpd/100t).
• Figure 3: Marginalized 68% and 95% credible intervals for $\mathcal{R}_{\mathrm{B}}$ and $\mathcal{R}_{\mathrm{ven}}$ after using uniform prior on $\frac{\mathcal{R}_{\mathrm{ven}}}{(1\,\mathrm{AU})^2} \in \mathcal{U} [-50, 50]$ and normal prior on $\mathcal{R}_{\mathrm{sun}} \in \mathcal{N} (25,2)$ with units of (cpd/100t). For this plot, we have used DE442s ephemeris. The marginalized 68% value for $\frac{\mathcal{R}_{\mathrm{ven}}}{(1\,\mathrm{AU})^2}$ is given by $0.18^{+0.09}_{-0.09}$ (cpd/100t).
• Figure 4: Marginalized 68% and 95% credible intervals for $\mathcal{R}_{\mathrm{sun}}$, $\mathcal{R}_{\mathrm{ven}}$ and $\mathcal{R}_{\mathrm{B}}$ after using uniform prior on $\frac{\mathcal{R}_{\mathrm{ven}}}{(1\,\mathrm{AU})^2} \in \mathcal{U} [-50, 50]$ and uniform prior on $\mathcal{R}_{sun} \in \mathcal{U} [0, 50]$ with units of (cpd/100t). For this plot, we have used DE442s ephemeris. The marginalized 68% value for $\mathcal{R}_{ven}\frac{\mathcal{R}_{\mathrm{ven}}}{(1\,\mathrm{AU})^2}$ is given by $0.16^{+0.09}_{-0.09}$ (cpd/100t).
• Figure 5: Marginalized 68% and 95% credible intervals for $\mathcal{R}_{\mathrm{B}}$ and $\mathcal{R}_{\mathrm{mar}}$ after using uniform prior on $\frac{\mathcal{R}_{\mathrm{mar}}}{(2\,\mathrm{AU})^2} \in \mathcal{U} [-13, 13]$ and normal prior on $\mathcal{R}_{\mathrm{sun}} \in \mathcal{N} (25,2)$ with units of (cpd/100t). For this plot, we have used DE442s ephemeris. The marginalized 68% value for $\frac{\mathcal{R}_{\mathrm{mar}}}{(2\,\mathrm{AU})^2}$ is given by $0.02^{+0.04}_{-0.04}$ (cpd/100t).